So after I found the bad driver chips and blown surface mount fuses,
and before I got the manual, (which does not have circuit diagrams anyway),
I replaced the drivers a few times and finally traced out the driver circuit back to the ink service station elevator motor.
(drivers are split over two driver chips for one motor)
About the time I ohmed out the shorted motor, I got the manual.
"If you replace the electronics module without replacing the bad motor,
the electronics module will blow again..." duh.
I figured out that the printer powers up and self tests each motor by sending a quick pulse and measuring the current drawn.
If the motor's coil resistance is wrong, the test stops and tells the operator that there is something wrong.
Rather than say "I think there is a failure in the ink service station motor, call for service", it says something like "err 917263876 or 63738292"
Very helpful HP. Have another Guiness.
What really happens is that the self test may or may not pass if your motor has a few shorted turns,
if it fails the test you get the error, If it passes the CPU continues to boot (several minutes) then the
motors are initiallized (homed) and when one of the motors is shorted, it blows the driver.
Because HP engineers spread the motor drives for a single motor over several driver chips, (presumabily to help with heat dissipation),
the bad motor blows the driver chip but the self test can't test the other motor that the driver runs so the diagnostic indicates the wrong motor is bad.
You go check the supposedly bad motor's resistance and it's OK.
It was only after I traced the lines from the driver to the elevator
motor that I found the problem. And then of course it was a special motor that could not be found anywhere.
